1.2.3 What about self-citations?
Whenever there is talk about citation analysis, the first concern always seems to be to exclude self-citations, i.e. authors citing their own work. It is generally believed that authors can inflate their citation records by excessive self-citations.
We will discuss this phenomenon in more detail in Section 9.2. However, I would like to categorically declare here that in my view the distrust of self-citations is completely misplaced. In most cases excluding self-citations is a huge waste of time and when done with ISI data has the potential to introduce more noise than it removes (see Section 14.2.2).
To provide an illustration of this I looked at a group of nearly 250 editors of 57 Management and Marketing journals between 1989 and 2009. The average number of self-citations for these academics was 4.3% at the time they became editor. The higher proportions occurred for the editors with fewer citations overall.
For those with less than 100 citations, the self-citation rate was on average 8%. For those at the opposite end of the spectrum, more than 1000 citations, the self-citation rate was only 2.4%. However, in neither case would be the number of self-citations be problematic enough to completely change the assessment of someone's publication record.
Only 30 out of the 242 editors had a self-citation rate of more than 10% (10 of which had a self-citation rate of 11%). Only five out of the 242 editors had a self-citation rate of more than 20% and three of those editors had less than 15 citations overall. In these cases self-citations are not the problem. They are often a legitimate way to acknowledge the academics previous research in the same field. What is a problem is a lack of non-self citations, i.e. the fact that other academics are not referring to these academics articles.